We study the contribution to the entanglement entropy of (2+1)-dimensional conformal field theories coming from a sharp corner in the entangling surface. This contribution is encoded in a function a(θ) of the corner opening angle, and was recently proposed as a measure of the degrees of freedom in the underlying CFT. We show that the ratio a(θ)/CT , where CT is the central charge in the stress tensor correlator, is an almost universal quantity for a broad class of theories including various higher-curvature holographic models, free scalars and fermions, and Wilson-Fisher fixed points of the O(N ) models with N = 1, 2, 3. Strikingly, the agreement between these different theories becomes exact in the limit θ → π, where the entangling surface approaches a smooth curve. We thus conjecture that the corresponding ratio is universal for general CFTs in three dimensions.Many interacting gapless quantum systems do not possess simple particle-like excitations, making it difficult to quantify their effective number of degrees of freedom (dof) at low-energy. Conformal field theories (CFTs) constitute an important example. For CFTs in two spacetime dimensions (2d), the Virasoro central charge is a good measure of the dof. It appears in many quantities, such as the thermal free energy and the entanglement entropy (EE), and decreases under renormalization group (RG) flow [1]. In higher dimensions, the concept of quantum entanglement is emerging as a fundamental diagnostic for such measures [2,3]. E.g., it was instrumental in finding an analogous RG monotone for 3d CFTs, with the EE of a disk-shaped region [4]. We shall study another measure of recent interest [5][6][7][8][9][10][11][12][13][14][15][16]: the coefficient capturing the contribution of sharp corners to spatial entanglement.In the context of quantum field theory, the EE is defined for a spatial region V as: S = −Tr (ρ V ln ρ V ), where ρ V is the reduced density matrix produced by integrating out the dof in the complementary region V . In the groundstate of a 3d CFT, the EE takes the form:where δ is a short-distance cutoff, e.g., the lattice spacing, and , a length scale associated with the size of V . The first, 'area law', term depends on the UV regulator and scales with the size of the boundary. The second one appears only when V has a sharp corner with opening angle θ ∈ [0, 2π), Fig. 1. Crucially, a(θ) is a regulator independent coefficient that characterizes the underlying CFT. It is positive and satisfies a(2π − θ) = a(θ) [5], and behaves as follows:in the limits of a nearly smooth entangling surface and a very sharp corner, respectively. It has been studied for a variety of systems: free scalars and fermions [5][6][7], interacting scalar theories via numerical simulations [8][9][10], Lifshitz quantum critical points [11], and holographicFIG. 1: a) An entangling region V of size with a corner; b) The holographic entangling surface γ for a region on the boundary of AdS4 with a corner.models [12]. The results suggest that a(θ) is an effective measure of the do...